304 ROMAN .VECHITECTURE. Part I. This, however, was not the course that the Romans pursued. What they did was to remove the pier altogether, and to substitute for it the pillar taken down from its pedestal. This of course was not effected at once, but was the result of many trials and expedients. One of the earliest of these is observed in the Ionic Temple of Con- cord before alluded to, in which a concealed arch is thrown from the head of each pillar, but above the entablature, so as to take the whole weight of the superstructure from off the cornice between the pillars. When once this was done, it was perceived that so deep an entablature ,lS!;lilltlii't 185. View in Courtyard of I'alace at .Si'alatro.' was no longer required, and that it might be either wholly omitted, as was sometimes done in the centre intercolumniation, or very much reduced. There is an old temple at Tnlavera in Spain, which is a o-ood example of the former expedient; and the Church of the Hojy Sepulchre, built by Constantine at Jerusalem, is a remarkable instance of the latter. There the architrave is cut off so as merely to foi-m a block over each of the i)illars, and the frieze and cornice only are carried across from one of these blocks to the other, while a bold arch is thrown from ]nllar to ])illar over these, so as to take any weight from off a member which has at last beconie a mere ornamental part of the style. In Diocletian's reign we find all these changes already introduced 1 It has recently become the fashion to spell the name Spalato or Spclato. The mode of writing it adopted in tliis work is that used by Adams, which has conseqnently become classical among architects.
